Amazon's multimodal foundation model on Bedrock, positioned for price-performance.
Specifications
- Provider
- Amazon
- Type
- Vendor / proprietary
- Modality
- Multimodal
- Category
- Multimodal model
- Context window
- 300K
- License
- Proprietary (Bedrock)
- Knowledge cutoff
- 2024
- Released
- December 3, 2024
What it was trained for
Multimodal understanding across text, image, and video with strong price-performance on AWS Bedrock.
Best for
- ▸Multimodal document and media analysis
- ▸Enterprise agents on AWS
- ▸Video and image understanding
- ▸Cost-sensitive production workloads
Capabilities
Image understandingVideo understandingTool use and function callingLong-context processing
Performance & positioning
Balances capability with low cost and fast inference within the Amazon Bedrock ecosystem.